肺应变对急性呼吸窘迫综合征模型犬呼吸力学的影响

Effect of lung strain on breathing mechanics in dogs with acute respiratory distress syndrome

目的探讨肺应变对急性呼吸窘迫综合征（ARDS）模型比格犬呼吸力学的影响。方法取24只成年健康雄性比格犬，采用静脉注射油酸0.18 m/g的方法制备中度ARDS模型。制模成功后按随机数字表法分为5组。肺保护性通气（LPV）组（n=4）潮气量（VT）设定为6~8 m/g；肺应变1.0、1.5、2.0、2.5组（S1.0、S1.5、S2.0、S2.5组）根据肺应变、呼气末正压引起的肺复张容积（VPEEP）和功能残气量（FRC）计算VT。5组均进行机械通气24 h或达到实验终点〔当去甲肾上腺素用量〉1.4 μg·kg-1·min-1仍不能使血压〉60 mmHg（1 mmHg=0.133 kPa）超过30 min视为实验终点〕。记录实验期间动物静态肺顺应性（Cst）、气道平台压（Pplat）和肺应力的变化。采用线性回归分析，拟合肺应变与Cst降变率、Pplat及肺应力的回归方程，探讨各指标之间的关系。结果LPV组VT为（7.1±0.5）m/g；随肺应变增加，VT逐渐升高。除S1.0组VT接近LPV组水平外〔为（7.3±1.8）m/g〕，S1.5、S2.0、S2.5组VT均较LPV组明显增加（m/g：13.3±5.5、18.7±5.4、20.1±7.4比7.1±0.5，均P〈0.05）；且相同肺应变时个体间VT差异较大。各组动物在实验终点时Cst均较制模前明显下降；随肺应变增加，Cst降变率呈升高趋势，S2.0、S2.5组Cst降变率明显大于S1.0、S1.5组〔（48.0±15.0）%、（54.4±9.5）%比（25.9±13.7）%、（38.6±8.1）%，均P〈0.05〕。各组动物成模时Pplat和肺应力均较制模前明显升高，并随通气时间延长呈升高趋势；S1.5组通气4 h Pplat和肺应力即明显高于LPV组〔Pplat（cmH2O，1 cmH2O=0.098 kPa）：26.2±2.3比20.2±4.2，肺应力（cmH2O）：20.5±2.0比16.6±2.5，均P〈0.05〕，且均随肺应变的增加而升高，直至实验终点。相关分析显示，肺应变与Cst降变率、通气4 h Pplat及肺应力均呈显著正相关（r值分别为0.716、0.660、0.539，均P〈0.05），提示存在较强的线性相关关系。拟合线性回归分析结果显示，肺应变每增加1，Cst降变率则增加19.0%〔95%可信区间（95%CI）=14.6~23.3，P=0.000〕，Pplat则升高10.8 cmH2O（95%CI=7.9~13.7，P=0.002），肺应力则升高7.4 cmH2O（95%CI=4.7~10.2，P=0.002）。结论在ARDS犬模型中，机械通气时肺应变越大，Pplat及肺应力越高，以肺应变2.0和2.5得到的VT可引起ARDS模型动物Cst进一步降低，肺应变大于1.5时Pplat及肺应力即显著增加，超过LPV的安全范围（分别为35 cmH2O和25 cmH2O），从而引起呼吸机相关性肺损伤（VILI）。

ObjectiveTo explore the effect of lung strain on breathing mechanics in dogs with acute respiratory distress syndrome （ARDS）.MethodsTwenty-four healthy male Beagle dogs were recruited to reproduce medium ARDS models with venous injection of 0.18 m/g oleic acid, and they were randomly assigned to five groups with random numbers table method. In lung protective ventilation （LPV） group （n = 4）, the ventilation was supported for 24 hours with tidal volume （VT） at 6-8 m/g, and in lung strain 1.0, 1.5, 2.0, 2.5 groups （S1.0, S1.5, S2.0, S2.5 groups）, the VT was calculated from lung strain, the volume recruitment by positive end expiratory pressure （VPEEP） and functional residual capacity （FRC）. Five groups were given mechanical ventilation for 24 hours or until reaching the end point of the experiment [when the dosage of norepinephrine was higher than 1.4 μg·kg-1·min-1, the blood pressure was still lower than 60 mmHg （1 mmHg = 0.133 kPa） for more than 30 minutes, which was regarded as the end point of the experiment]. Static lung compliance （Cst）, airway plateau pressure （Pplat） and lung stress during the experiment were recorded. Linear regression analysis was used to fit the regression equations of lung strain and Cst descending rate, Pplat and lung stress for analyzing their relationships.ResultsThe VT of group LPV was （7.1±0.5） m/g. With the increase of lung strain, VT was gradually increased. VT of group S1.0 [（7.3±1.8） m/g] was similar to group LPV. VT of groups S1.5, S2.0, S2.5 was significantly higher than that of group LPV （m/g： 13.3±5.5, 18.7±5.4, 20.1±7.4 vs. 7.1±0.5, all P 〈 0.05）. Moreover, under the same lung strain, the difference in VT among individuals was large. The Cst of each group was decreased significantly at the end of the experiment as compared with that before model reproduction. With the increase of lung strain, the rate of Cst descending was increased, Cst dropped more significantly in groups S2.0 and S2.5 than that in groups S1.0 and S1.5 [（48.0±15.0）%, （54.4±9.5）% vs. （25.9±13.7）%, （38.6±8.1）%, all P 〈 0.05]. Pplat and pulmonary stress at model reproduction in all groups were significantly higher than those before model reproduction, and they increased with the prolongation of ventilation time. Pplat and lung stress at 4 hours of ventilation in group S1.5 were significantly higher than those in group LPV [Pplat （cmH2O, 1 cmH2O = 0.098 kPa）： 26.2±2.3 vs. 20.2±4.2, lung stress （cmH2O）： 20.5±2.0 vs. 16.6±2.5, both P 〈 0.05], and they increased with lung strain increasing till to the end of experiment. It was shown by correlation analysis that lung strain was positively related with Cst descending rate, Pplat and lung stress at 4 hours of ventilation （r value was 0.716, 0.660, 0.539, respectively, all P 〈 0.05）, which indicated a strong linear correlation. It was shown by fitting linear regression analysis that when lung strain increased by 1, Cst descending rate increased by 19.0% [95% confidence interval （95%CI） = 14.6-23.3, P = 0.000], Pplat increased by 10.8 cmH2O （95%CI = 7.9-13.7, P = 0.002）, and the lung stress increased by 7.4 cmH2O （95%CI = 4.7-10.2, P = 0.002）.ConclusionIn animal ARDS models, the larger the lung strain, the higher the Pplat and lung stress during mechanical ventilation, VT originated for lung strain 2.0 and 2.5 may further reduce Cst in ARDS models, when lung strain over 1.5, Pplat and lung stress increased significantly, which exceeded the safe range of LPV （35 cmH2O and 25 cmH2O, respectively）, and further aggravated ventilator induced lung injury （VILI）.